Online etymological dictionary says:
O.E. spelt, perhaps an early borrowing from L.L. spelta "spelt" (c.400, noted as a foreign word), which is perhaps ultimately from PIE root *spel- "to split, to break off" (probably in reference to the splitting of its husks in threshing), which is related to the root of flint. The word had little currency in English, and its history is discontinuous. Widespread in Romanic languages (cf. It. spelta, Sp. espelta, O.Fr. spelte, Mod.Fr. épeautre). The word also is widespread in Germanic (cf. O.H.G. spelta, Ger. Spelt), and a Germanic language is perhaps the source of the Late Latin word.
So no need to look back to Indo-European e/o variants. Kluge's German etymological dictionary says it's one of the earliest Germanic borrowings into Latin.
French é is not always from es. Past participle é is Latin at-.
I don't think the lt and the tr are a metathesis. The word first appears in French as espiaute, 1256. Forms without r are found as late as the XVIII century. (Albert Dauzat et al, Nouveau Dictionnaire Etymologique, which either doesn't seem to know qbout the OF spelte that etymonline mentions.) You expect an l to disappear in that environment; cf. faute from Latin fallita, faucon from Latin falcon-. I don't know why the -re developed.
Offhand, the only r/l cognates I can think of are English words where an r had dissimilated. Purple from purpura; marble from marbre.
If they wwre from the same IE root rather than reflecting a borrowing, you wouldn't expect a t in both, on account of Grimm's law. (in mlst cases you wouldn't expect IE p to remain in Germanic either, but Grimm's law doesn't operqte on consonants immediately after s. Spuo/spit; stella/star.
Edited by: VinnyD to add last bit.
